Visible Light Image-Based Method for Sugar Content Classification of Citrus.

Area of Science:

• Agricultural Science
• Image Processing
• Food Science

Background:

• Predicting citrus fruit sugar content is crucial for quality control and consumer satisfaction.
• Current methods for determining sugar content can be destructive or labor-intensive.
• Developing a nondestructive, accurate method for sugar content prediction is highly desirable.

Purpose of the Study:

• To develop and validate an algorithm for predicting citrus fruit sugar content using visible light imaging.
• To explore the correlation between color image parameters and actual sugar content in citrus fruits.
• To establish a nondestructive classification method for citrus fruit sweetness.

Main Methods:

• Visible light imaging was employed on citrus fruits from Mie Prefecture, Japan.
• Correlation analysis, using the coefficient of determination, was applied for image segmentation.
• An addition algorithm connected selected image parameters, and the dummy variable method was used for prediction.

Main Results:

• Accurate image segmentation was achieved through correlation analysis.
• A significant correlation was identified between citrus fruit sugar content and specific color image parameters.
• The developed algorithm successfully predicted sugar content, with smaller, orange fruits indicating higher sweetness.

Conclusions:

• Visible light imaging offers a viable, nondestructive method for predicting citrus fruit sugar content.
• The algorithm enables classification of sugar content without additional light sources.
• This technique has potential applications in citrus quality assessment and management.